Pancreatic adenocarcinoma (PAAD) is a lethal disease with a poor prognosis. Genes involved in acute pancreatitis (AP) or chronic pancreatitis (CP) might be important for PAAD development. This study sought to identify potential PAAD diagnosis markers and to establish a PAAD prognosis prediction model based on AP- and CP-related genes. The significantly differentially expressed genes in both AP or CP and PAAD were obtained by a bioinformatics analysis. A risk-score model for predicting survival was constructed based on The Cancer Genome Atlas (TCGA) data and validated using an International Cancer Genome Consortium (ICGC) cohort. Protein expression and the effects of the genes in the risk models were validated by immunohistochemistry, or Cell Counting Kit-8 (CCK-8) and transwell assays. The study sample data included six AP tissue samples and five normal pancreatic tissue samples, six CP tissue samples and six normal pancreatic tissue samples from the Gene Expression Omnibus (GEO) expression profiling microarrays GSE109227 and GSE41418 data sets, respectively, and fragments per kilobase per million mapped fragments (FPKM) data from four normal controls and 150 PAAD cases from TCGA database, and 182 cancer patient samples with complete survival prognostic data from the ICGC database. In total, 508 significantly differentially expressed genes were found in both AP or CP and PAAD. Trefoil factor 2 (TFF2), tubulointerstitial nephritis antigen (TINAG), trefoil factor 1 (TFF1), aquaporin 5 (AQP5), SAM pointed domain containing ETS transcription factor (SPDEF), anterior gradient protein 2 (AGR2), apolipoprotein B messenger RNA editing enzyme catalytic subunit 1 (APOBEC1), kallikrein-related peptidase 6 (KLK6), dopa decarboxylase (DDC), mucin 13 (MUC13), claudin 18 (CLDN18), annexin A10 (ANXA10), and tetraspanin 1 (TSPAN1) were found to be present in PAAD and had the largest fold change. A risk-score model, comprising 19 genes, was constructed for prognostic prediction. A high-risk score indicated a poor prognosis. TINAG, DDC, SPDEF, and APOBEC1 proteins were increased in PAAD, while TINAG and DDC were correlated with the pathologic grade. Decreased TINAG, APOBEC1, transmembrane protein 94 (TMEM94), and kelch like family member 36 (KLHL36) expression inhibited PAAD cell proliferation, while decreased SPDEF, TMEM94, and KLHL36 expression significantly inhibited PAAD cell migration. The AP and CP co-related genes were significantly correlated with PAAD. TINAG, DDC, SPDEF, and APOBEC1 could serve as new PAAD predictors. The risk model developed in this study could be used to predict the prognosis of PAAD patients.